linux-kernel - Re: [PATCH] hung_task: recover hung task warnings in next check interval

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CALOAHbCe9J0pOCW03dW+C4NK__amTKttAs=eNHXwvPPf5Lpwhw@mail.gmail.com>
Date:   Thu, 20 Jun 2019 18:10:06 +0800
From:   Yafang Shao <laoar.shao@...il.com>
To:     Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>
Cc:     Andrew Morton <akpm@...ux-foundation.org>,
        Dmitry Vyukov <dvyukov@...gle.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] hung_task: recover hung task warnings in next check interval

On Thu, Jun 20, 2019 at 6:03 PM Tetsuo Handa
<penguin-kernel@...ove.sakura.ne.jp> wrote:
>
> On 2019/06/20 14:55, Yafang Shao wrote:
> > When sys_hung_task_warnings reaches 0, the hang task messages will not
> > be reported any more.
>
> It is a common mistake that sys_hung_task_warnings is already 0 when
> a real problem which should be reported occurred.
>
> >
> > If the user want to get more hung task messages, he must reset
> > kernel.hung_task_warnings to a postive integer or -1 with sysctl.
>
> People are setting sys_hung_task_warnings to -1 in order to make sure
> that the messages are printed.
>
> > This is not a good way for the user.
>
> But I don't think we should reset automatically.
>
> > We'd better reset hung task warnings in the kernel, and then the user
> > don't need to pay attention to this value.
>
> I suggest changing the default value of sys_hung_task_warnings to -1.
>

Yes, that's what we have did now.

> >
> > With this patch, hung task warnings will be reset with
> > sys_hung_task_warnings setting in evenry check interval.
>
> Since it is uncommon that the messages are printed for more than 10
> times for one check_hung_uninterruptible_tasks() call, this patch is
> effectively changing to always print the messages (in other words,
> setting -1).

If sys_hung_task_warnings can't be recovered, does it make sense to exist?
In which case do we need this setting ?

Btw, why the default value of this setting is 10, instead of -1 ?

Thanks
Yafang